One of the most popular treatments for autism spectrum disorders (ASDs) is a diet free of gluten and casein (the GFCF diet). Whether or not the theory on which the diet is based has merit, many parents report a decrease in behavioral problems when it is followed and report an increase in behavioral problems with exposure to gluten or casein. As with other interventions, there appear to be children who respond well to this treatment and others for whom it has little or no effect. Because families choosing or rejecting this diet are likely to represent groups that differ in ways difficult to assess, we propose that the effects of the diet are best studied in a cohort of children in which all receive the diet. The children will then be tested for their response to challenges with gluten and casein. All children who participate will be well-characterized before the diet is instituted and will receive intensive early behavioral intervention in the same program throughout the 18 weeks of the study. A nutritionist will assist families with compliance and with planning diets that are nutritionally complete, and nutritional status will be monitored throughout the study. Participants will receive challenges at school in the form of specially prepared and individually administered snacks. Parents, teachers, and the research team will be blind to the contents of the snacks. Teachers and parents will monitor three kinds of behaviors: a) behaviors related to ASDs [e.g., stereotyped behaviors, social reciprocity], b) behaviors not specific to ASDs [e.g., activity level, sleep pattern, appetite], and c) constitutional measures [e.g., stooling pattern, heart rate]. The first goal of the study is to determine whether challenges elicit measurable behavioral change. The second is to determine whether phenotypic characteristics of the participants identify responders and non-responders. The third goal is to determine whether any of six proposed genetic susceptibility factors for ASDs are related to outcome of the GFCF diet treatment.